Voice Cloning Policy

Effective date: 01.09.2026

Controller: ImStocker LLC ("TutDub", "we", "us", "our")

This Policy describes how TutDub's voice-cloning feature works, the consent and rights we require, the safeguards we apply, and the prohibited uses. It supplements the Terms of Service and Privacy Policy.

1. What "Voice Cloning" Means Here

TutDub can replace the original speech in a video/audio with a translated version spoken in the same voice. To do this, the Service:

Transcribes the original speech (ASR).
Translates the transcript with a language model.
Selects a short reference audio slice from the speaker's own voice as it appears in the uploaded video/audio (the longest detected speech segment, capped at ~10 seconds).
Re-synthesizes the translated text in that voice using a zero-shot voice-cloning text-to-speech model, conditioned on the reference slice.
Mixes the new voice back into the video/audio (optionally preserving background audio and adapting video timing).

The cloned voice is therefore derived entirely from the video/audio you upload. TutDub does not ask you to upload a separate voice sample, and there is no library of celebrity or third-party voices.

2. Consent and Rights Confirmation

Before a dubbing job using voice cloning is accepted, the application requires you to explicitly:

Agree to clone the voice from the submitted footage (mandatory consent checkbox), and
Confirm that you have the rights to use the content and the voices in it (author/rights confirmation checkbox).

This consent is required for each video/audio you upload.

Voice cloning is integral to how TutDub produces dubbed videos/audios: every dubbing job re-synthesizes the translated speech in the speaker's own voice taken from the uploaded footage. Cloning cannot be turned off for a job, so the consent and rights checkboxes above are a required step before any video/audio is processed.

Revocability You may withdraw your consent at any time by deleting the video/audio/Output from your account. Upon withdrawal, we will cease further cloning for that content and delete the associated voice files as described in Section 4. Withdrawal of consent does not affect the lawfulness of processing based on consent before its withdrawal.

Multiple Speakers If your video/audio contains the voices of multiple identifiable individuals, you must obtain informed consent from each individual whose voice is cloned before uploading. You are solely responsible for securing this consent. "Identifiable individuals" includes all voices that can be recognized, not just the primary speaker.

By submitting Content for dubbing, you represent and warrant that:

You have the necessary rights, licenses, and consents to clone the voices present in the video/audio; and
Cloning those voices for the intended Output does not infringe any third party's personality, privacy, copyright, or other rights.

If the video/audio contains voices of other identifiable individuals, it is your responsibility to ensure you have their informed consent or another lawful basis before enabling cloning.

3. What We Do NOT Do

We do not create or retain a persistent, reusable "voice model," voiceprint, or biometric profile of any speaker. The reference slice is used only for the specific job and is deleted together with that job's working files after the Output is produced and stored.
We do not maintain a catalog of voices, and we do not offer voice cloning of anyone who does not appear in the video/audio you provide.
We do not use uploaded voices to train our models or to improve the Service beyond the single job, and we do not share voice data with advertisers.
We do not use cloning to generate speech for individuals who have not appeared in your uploaded Content.

4. How Voice Data Is Processed and Retained

The reference audio and all intermediate voice-synthesis files are created in a per-job working directory on our processing worker and are deleted automatically after the job completes and its Output is uploaded to storage.
The resulting dubbed video and audio (which contain the synthesized voice) are stored in storage and follow the standard 96-hour retention window measured from the moment the final processing job completes (i.e., after all selected language versions have been generated and the Output is ready for download).
After the 96-hour retention window expires, the files are permanently deleted from storage. Database metadata (e.g., job ID, timestamp, languages selected) is retained as described in the Privacy Policy for billing, analytics, and service improvement purposes.
Voice data is processed on our own or leased compute infrastructure; it is not sent to third parties except the storage provider that hosts the resulting files.

5. Prohibited Uses

You may not use voice cloning to:

Impersonate any person without their consent, or depict them saying things they did not say, for deception, fraud, defamation, harassment, or harm.
Create misleading "deepfake" content intended to deceive viewers about the origin, endorsement, or statements of any individual or organization.
Clone the voice of a minor without appropriate parental/guardian consent.
Use voice cloning for commercial exploitation (e.g., advertising, product endorsements, audiobooks, or any paid content) without explicit consent from the voice owner.
Create content that misrepresents a political figure's statements, positions, or endorsements.
Generate voice clones for use in sexually explicit, pornographic, or otherwise obscene content.
Infringe copyright, rights of publicity/personality, or privacy, or otherwise violate the Terms of Service or applicable law.
Facilitate any of the above activities, or assist others in doing so.

We may refuse, suspend, or terminate processing and accounts that engage in or facilitate such misuse, and we may report unlawful content where required.

6. Quality, Risks, and Disclosure

6.1 Quality and Review

Cloning is automated and may produce imperfect results (accent, prosody, emotional tone, or intelligibility). You are responsible for reviewing Output before publishing or relying on it.

6.2 Disclosure of Synthetic Content

TutDub produces AI-generated voice outputs. When you publish or distribute content created using TutDub, you must clearly disclose that the voice has been artificially generated or synthesized.

Examples of appropriate disclosure:

"We used #TutDub for voice translation "
"This video/audio contains AI-generated voice cloning #TutDub."
"Voice synthesized using #TutDub."
"The voice in this video/audio is artificially generated #TutDub."

We recommend placing this disclosure prominently, such as in the video/audio description, on-screen text, or audio introduction or any other context where disclosure is reasonably expected. This practice helps maintain trust with your audience and reduces the risk of misunderstanding or misuse.

By using TutDub, you agree to comply with this disclosure requirement.

6.3 Prohibited Use and Reporting

Synthetic speech can be misused. We apply the consent and rights controls in Section 2 and the deletion practices in Section 4 to mitigate risks. No technical control is perfect. Report suspected abuse to support@tutdub.com.

7. Your Controls and Rights

Decline to submit content containing voices you do not have the rights to clone. Because voice cloning is an inseparable part of the Service, the only way to avoid cloning a given person's voice is not to upload video/audio that contains it.
Delete videos/Output from your account at any time, which removes the stored files from storage.
Request erasure of any retained voice-related metadata via support@tutdub.com; note that, by design, per-job voice working files are already deleted after processing.
Withdraw consent at any time by deleting the video/audio/Output from your account. Upon withdrawal, we will cease further cloning for that content and delete the associated files as described in Section 4.
Where voice is considered biometric or sensitive under applicable law, we rely on your explicit consent (captured at upload) and honor its withdrawal by ceasing further cloning for affected content.

8. Legal Classification Note

In some jurisdictions, a recording of a person's voice used to identify them may be treated as biometric or otherwise special-category data. We treat voice data with corresponding care: limited purpose (only to produce your Output), no persistent voiceprints, and deletion by default. This does not constitute a legal determination; applicable classifications depend on your jurisdiction and use case.

9. Changes to This Policy

We may update this Policy from time to time. If we make material changes, we will notify you by email or through the Service. Your continued use of TutDub after the changes take effect constitutes your acceptance of the revised Policy.

10. Contact

Questions or abuse reports regarding voice cloning: support@tutdub.com or via the feedback form at https://tutdub.com/en/feedback.

ImStocker LLC

Address: Apt. 191, 17 9-ya Podlesnaya St., Izhevsk, Udmurt Republic, 426054, Russia

Email: support@tutdub.com